Risk factors for necrotizing enterocolitis totalis: a case-control study

The objective of this study was to compare demographic and clinical events in three groups of preterm neonates: those with necrotizing enterocolitis totalis (NEC-T), those with NEC non-totalis (NEC non-T) and in preterm patients without NEC.

This retrospective case-control study was conducted at Yale New Haven Children's Hospital using patient data from January 1991 to December 2007. Study patients were less than 36 weeks of gestational age (GA) at birth, without gastrointestinal (GI) malformations. Cases (NEC-T) were diagnosed at operation or at autopsy with observation of >80% necrosis of the GI tract. Two control groups were assigned: Group 1 or NEC non-T and Group II or Non-NEC. Two to four controls per case were matched to cases by GA at birth±2 weeks. Demographic and clinical data for the day of diagnosis and retrospectively up to 7 days preceding diagnosis were recorded for those with NEC-T and NEC. Group II controls were matched for date of birth and day of life, in addition to GA at birth.

A total of 14 075 patients were admitted to the Newborn Special Care Unit during the study interval. Overall 328 patients (2.3%) developed NEC≥Bell's Stage II; 39 patients met inclusion criteria for NEC-T case status; 148 NEC non-T and 110 non-NEC controls were assigned. In the comparison of NEC T and NEC non-T neonates, use of breast milk was associated with decreased risk of NEC-T, adjusted odds ratio (OR)=0.26, 95% confidence interval (CI) of OR=0.08-0.085, P=0.03. When NEC T and non-NEC patients were compared, having reached full-enteral feeds before the date of diagnosis of the matched case (adjusted OR=28.5, 95% CI of OR=2.7-299, P=0.005) and use of breast milk (adjusted OR=0.09, 95% CI of OR=0.02-0.56, P=0.01) were significantly different between the two groups.

Breast milk usage was significantly associated with decreased occurrence of NEC-T in our comparison of NEC-T, NEC non-T and non-NEC patients. Although there were some differences, the majority of demographic and clinical variables assessed were not shown to be significantly different between cases and controls. This highlights the need for more biological data in assessing risk of developing NEC-T.